The Emergency Surgery Frailty Index (EmSFI): development and internal validation of a novel simple bedside risk score for elderly patients undergoing emergency surgery

Background Frailty assessment has acquired an increasing importance in recent years and it has been demonstrated that this vulnerable profile predisposes elderly patients to a worse outcome after surgery. Therefore, it becomes paramount to perform an accurate stratification of surgical risk in elderly undergoing emergency surgery. Study design 1024 patients older than 65 years who required urgent surgical procedures were prospectively recruited from 38 Italian centers participating to the multicentric FRAILESEL (Frailty and Emergency Surgery in the Elderly) study, between December 2016 and May 2017. A univariate analysis was carried out, with the purpose of developing a frailty index in emergency surgery called "EmSFI". Receiver operating characteristic curve analysis was then performed to test the accuracy of our predictive score. Results 784 elderly patients were consecutively enrolled, constituting the development set and results were validated considering further 240 consecutive patients undergoing colorectal surgical procedures. A logistic regression analysis was performed identifying different EmSFI risk classes. The model exhibited good accuracy as regard to mortality for both the development set (AUC = 0.731 [95% CI 0.654-0.772]; HL test chi(2) = 6.780; p = 0.238) and the validation set (AUC = 0.762 [95% CI 0.682-0.842]; HL test chi(2) = 7.238; p = 0.299). As concern morbidity, our model showed a moderate accuracy in the development group, whereas a poor discrimination ability was observed in the validation cohort. Conclusions The validated EmSFI represents a reliable and time-sparing tool, despite its discriminative value decreased regarding complications. Thus, further studies are needed to investigate specifically surgical settings, validating the EmSFI prognostic role in assessing the procedure-related morbidity risk.